Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    trophy

    np_max-act · gemini-2.0-flash

    This neuron activates on the substring “trophy,” spotting any token that contains “trophy” (e.g., trophy, trophies, atrophy).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    references to trophies.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     setters
    -0.08
    Scene
    -0.08
    Current
    -0.07
    .Emit
    -0.07
    028
    -0.07
    856
    -0.07
    232
    -0.07
     Verizon
    -0.07
     beams
    -0.07
    Binder
    -0.07
    POSITIVE LOGITS
    rophy
    0.10
     Trophy
    0.10
     trophies
    0.09
     trophy
    0.09
    roph
    0.08
    pro
    0.07
    лиш
    0.07
     Troll
    0.07
    생님
    0.07
    ΟΦ
    0.06
    Act Density 0.004%

    No Known Activations